GL07

Until recently, it was commonly written that 2 billion people are latently infected with Mycobacterium tuberculosis and that up to 10% of these could ‘reactivate’, resulting in ‘active’ TB disease.  This led to an epidemiologic paradigm where a small number of new infections cause the 10 million new cases after a ‘long and variable’ incubation period.  This also stimulated fundamental research on induction and resuscitation of bacteria to/from a state of dormancy.  A review of the historical literature has revealed multiple lines of evidence that disease most frequently occurs in the first year after infection.  The annual risk of infection might be ten-fold higher than currently estimated, but the long-term risk much lower, in part because those who do not progress to infection often self-clear within 5 years.  Some of the confusion may result from the variable use of the word ‘latent’ over two centuries. Pierre-Charles-Alexandre Louis and Rene Laennec used latent in the 1820s to describe a host with pathology but no clinical manifestations. Latent was only applied to the bacterium in the early 20^th century when latent bacilli were presented as a hypothesized explanation for bacterial detection in the absence of classic pathology.  A review of the terminology applied to describe infection and disease indicates that adjectives can be misleading, especially if applied by some authors to hosts and by others to bacteria. We propose that tests we use for M. tuberculosis infection be recognized as indirect, that the concept of infection without symptoms or without a positive sputum culture should be called tuberculous infection (TBI), and that the term for the disease should simply be tuberculosis (TB).